Abstract-This paper develops a new cooperative jamming protocol, termed accumulate-and-jam (AnJ), to improve physical layer security in wireless communications. Specifically, a fullduplex (FD) friendly jammer is deployed to secure the direct communication between source and destination in the presence of a passive eavesdropper. We consider the friendly jammer as an energy-constrained node without embedded power supply but with an energy harvesting unit and rechargeable energy storage; it can thus harvest energy from the radio frequency (RF) signals transmitted by the source, accumulate the energy in its battery, and then use this energy to perform cooperative jamming. In the proposed AnJ protocol, based on the energy status of the jammer and the channel state of source-destination link, the system operates in either dedicated energy harvesting (DEH) or opportunistic energy harvesting (OEH) mode. In DEH mode, the source sends dedicated energy-bearing signals and the jammer performs energy harvesting. In OEH mode, the source transmits an information-bearing signal to the destination. Meanwhile, using the harvested energy, the wireless-powered jammer transmits a jamming signal to confound the eavesdropper. Thanks to the FD capability, the jammer also harvests energy from the informationbearing signal that it overhears from the source. We study the complex energy accumulation and consumption procedure at the jammer by considering a practical finite-capacity energy storage, of which the long-term stationary distribution is characterized through applying a discrete-state Markov Chain. An alternative energy storage with infinite capacity is also studied to serve as an upper bound. We further derive closed-form expressions for two secrecy metrics, i.e., secrecy outage probability and probability of positive secrecy capacity. In addition, the impact of imperfect channel state information on the performance of our proposed protocol is also investigated. Numerical results validate all theoretical analyses and reveal the merits of the proposed AnJ protocol over its half-duplex counterpart.Index Terms-Cooperative jamming, full-duplex, imperfect channel state information, physical layer security, wireless energy harvesting.

Sensitive and spatial exploration of the metabolism of tumors at the metabolome level is highly challenging. In this study, we developed an in situ metabolomics method based on ambient mass spectrometry imaging using air flow-assisted desorption electrospray ionization (AFADESI), which can spatially explore the alteration of global metabolites in tissues with high sensitivity. Using this method, we discovered potential histopathological diagnosis biomarkers (including lipids, amino acids, choline, peptides, and carnitine) from 52 postoperative lung cancer tissue samples and then subsequently used these biomarkers to generate images for rapid and label-free histopathological diagnosis. These biomarkers were validated with a sensitivity and a specificity of 93.5% and 100%, respectively. Moreover, a single imaging analysis of a cryosection that visualized all these biomarkers, taking tens of minutes, revealed the type and subtype of the cancer. This method could potentially be used as a molecular pathological tool for rapid clinical lung cancer diagnosis and immediate image-guided surgery.

An ensemble consists of multiple learners and can achieve a better generalisation performance than a single learner. Genetic programming (GP) has been applied to construct ensembles using different strategies such as bagging and boosting. However, no GP-based ensemble methods focus on dealing with image classification, which is a challenging task in computer vision and machine learning. This paper proposes an automated ensemble learning framework using GP (EGP) for image classification. The new method integrates feature learning, classification function selection, classifier training, and combination into a single program tree. To achieve this, a novel program structure, a new function set and a new terminal set are developed in EGP. The performance of EGP is examined on nine different image classification data sets of varying difficulty and compared with a large number of commonly used methods including recently published methods. The results demonstrate that EGP achieves better performance than most competitive methods. Further analysis reveals that EGP evolves good ensembles simultaneously balancing diversity and accuracy. To the best of our knowledge, this study is the first work using GP to automatically generate ensembles for image classification.

Artificial intelligence (AI) emphasises the creation of intelligent machines/systems that function like humans. AI has been applied to many real-world applications. Machine learning is a branch of AI based on the idea that systems can learn from data, identify hidden patterns, and make decisions with little/minimal human intervention. Evolutionary computation is an umbrella of population-based intelligent/learning algorithms inspired by nature, where New Zealand has a good international reputation. This paper provides a review on evolutionary machine learning, i.e., evolutionary computation techniques for major machine learning tasks such as classification, regression and clustering, and emerging topics including combinatorial optimisation, computer vision, deep learning, transfer learning, and ensemble learning. The paper also provides a brief review of evolutionary learning applications, such as supply chain and manufacturing for milk/dairy, wine and seafood industries, which are important to New Zealand. Finally, the paper presents current issues with future perspectives in evolutionary machine learning.

Feature extraction is an essential process to image classification. Existing feature extraction methods can extract important and discriminative image features but often require domain expert and human intervention. Genetic Programming (GP) can automatically extract features which are more adaptive to different image classification tasks. However, the majority GPbased methods only extract relatively simple features of one type i.e. local or global, which are not effective and efficient for complex image classification. In this paper, a new GP method (GP-GLF) is proposed to achieve automatically and simultaneously global and local feature extraction to image classification. To extract discriminative image features, several effective and well-known feature extraction methods, such as HOG, SIFT and LBP, are employed as GP functions in global and local scenarios. A novel program structure is developed to allow GP-GLF to evolve descriptors that can synthesise feature vectors from the input image and the automatically detected regions using these functions. The performance of the proposed method is evaluated on four different image classification data sets of varying difficulty and compared with seven GP based methods and a set of non-GP methods. Experimental results show that the proposed method achieves significantly better or similar performance than almost all the peer methods. Further analysis on the evolved programs shows the good interpretability of the GP-GLF method.